Replace DB_URL with DATABASE_URL

This way it is shared with diesel, which simplifies a lot the setup

Also fixes a few issues in the documentation, that are not directly related
This commit is contained in:
Bat 2018-10-06 19:17:36 +01:00
parent 7a64005ca9
commit 9cc795d8be
8 changed files with 47 additions and 43 deletions

View file

@ -12,7 +12,7 @@ RUN apt-get update && apt-get install -y --no-install-recommends \
libssl-dev
WORKDIR /app
COPY Cargo.toml Cargo.lock ./
RUN cargo install diesel_cli --no-default-features --features postgres --version '=1.2.0'
RUN cargo install diesel_cli --no-default-features --features postgres --version '=1.3.0'
COPY . .
RUN cargo install --force
RUN cargo install --path plume-cli --force

View file

@ -6,9 +6,7 @@ starting the app with `cargo run` or write them in a `.env` file to have automat
Here are the variables that Plume uses:
- `BASE_URL`: the domain name, or IP and port on which Plume is listening. It is used in all federation-related code.
- `DB_URL`: the URL of the PostgreSQL database, used by Plume (`postgres://plume:plume@localhost/plume` by default).
- `POSTGRES_USER`: if you just want to use a different PostgreSQL user name, and keep the rest of the default URL.
- `POSTGRES_PASSWORD`: same as `POSTGRES_USER`, but for the password.
- `DATABASE_URL`: the URL of the PostgreSQL database, used by Plume (`postgres://plume:plume@localhost/plume` by default with PostgreSQL, `plume.db` with SQlite).
- `USE_HTTPS`: if it is `0`, federation and medias will be using HTTP by default (`1` by default).
- `ROCKET_ADDRESS`: the adress on which Plume should listen (`0.0.0.0` by default).
- `ROCKET_PORT`: the port on which Plume should listen ([`7878` by default](https://twitter.com/ag_dubs/status/852559264510070784))

View file

@ -133,6 +133,31 @@ export FEATURES=postgres
export FEATURES=sqlite
```
## Configuring Plume
Before starting Plume, you'll need to create a configuration file, called `.env`. Here is a sample of what you should put inside.
```bash
# The address of the database
# (replace USER, PASSWORD, PORT and DATABASE_NAME with your values)
#
# If you are using SQlite, use the path of the database file (`plume.db` for instance)
DATABASE_URL=postgres://USER:PASSWORD@IP:PORT/DATABASE_NAME
# For PostgreSQL: migrations/postgres
# For SQlite: migrations/sqlite
MIGRATION_DIR=migrations/postgres
# The domain on which your instance will be available
BASE_URL=plu.me
# Secret key used for private cookies and CSRF protection
# You can generate one with `openssl rand -base64 32`
ROCKET_SECRET_KEY=
```
For more information about what you can put in your `.env`, see [the documentation about environment variables](ENV-VARS.md).
## Running migrations
Migrations are scripts used to update the database. They are run by a tool called Diesel, which can be installed with:
@ -144,11 +169,7 @@ cargo install diesel_cli --no-default-features --features $FEATURES --version '=
To run the migrations, you can do:
```bash
# For a PostgreSQL database
diesel migration run --database-url postgres://USER:PASSWORD@IP:PORT/DATABASE_NAME
# For a SQlite database
diesel migration run --database-url ./plume.sqlite3
diesel migration run
```
Migrations should be run before using Plume or the `plm` CLI tool, and after each update.
@ -163,25 +184,6 @@ cargo install --no-default-features --features $FEATURES
cargo install --no-default-features --features $FEATURES --path plume-cli
```
Before starting Plume, you'll need to create a configuration file, called `.env`. Here is a sample of what you should put inside.
```bash
# The address of the database
# (replace USER, PASSWORD, PORT and DATABASE_NAME with your values)
#
# If you are using SQlite, use the path of the database file
DB_URL=postgres://USER:PASSWORD@IP:PORT/DATABASE_NAME
# The domain on which your instance will be available
BASE_URL=plu.me
# Secret key used for private cookies and CSRF protection
# You can generate one with `openssl rand -base64 32`
ROCKET_SECRET_KEY=
```
For more information about what you can put in your `.env`, see [the documentation about environment variables](ENV-VARS.md).
After that, you'll need to setup your instance, and the admin's account.
```
@ -224,8 +226,8 @@ docker-compose up -d postgres
docker-compose run --rm plume diesel migration run
# Setup your instance
docker-compose run --rm plume plume instance new
docker-compose run --rm plume plume users new --admin
docker-compose run --rm plume plm instance new
docker-compose run --rm plume plm users new --admin
# Launch your instance for good
docker-compose up -d
@ -489,8 +491,7 @@ exit 0
## Caveats:
- Pgbouncer is not supported yet (named transactions are used).
- Rust nightly is a moving target, dependancies can break and sometimes you need to check a few versions to find the one working (run `rustup override set nightly-2018-05-15` or `rustup override set nightly-2018-05-31` in the Plume directory if you have issues during the compilation)
- Rust nightly 2018-06-28 is known to be failing to compile diesel 1.3.2
- Rust nightly is a moving target, dependancies can break and sometimes you need to check a few versions to find the one working (run `rustup override set nightly-2018-07-17` in the Plume directory if you have issues during the compilation)
## Acknowledgements

View file

@ -7,7 +7,7 @@ git pull origin master
cargo install --force && cargo install --path plume-cli --force
# Run the migrations
diesel migration run --database-url 'YOUR_DB_URL'
diesel migration run
# If you are using sysvinit
sudo service plume restart

View file

@ -3,10 +3,8 @@ BASE_URL=yourdomain.com
ROCKET_SECRET_KEY=randomstringhere
# you can safely leave those defaults
POSTGRES_USER=plume
POSTGRES_PASSWORD=plume
DB_URL=postgres://plume:plume@postgres:5432/plume
DATABASE_URL=postgres://plume:plume@postgres:5432/plume
MIGRATION_DIR=postgres
USE_HTTPS=1
ROCKET_ADDRESS=0.0.0.0
ROCKET_PORT=7878

View file

@ -7,7 +7,7 @@ extern crate rpassword;
use clap::App;
use diesel::Connection;
use std::io::{self, prelude::*};
use plume_models::{DB_URL, Connection as Conn};
use plume_models::{DATABASE_URL, Connection as Conn};
mod instance;
mod users;
@ -22,7 +22,7 @@ fn main() {
let matches = app.clone().get_matches();
dotenv::dotenv().ok();
let conn = Conn::establish(DB_URL.as_str());
let conn = Conn::establish(DATABASE_URL.as_str());
match matches.subcommand() {
("instance", Some(args)) => instance::run(args, &conn.expect("Couldn't connect to the database.")),

View file

@ -191,12 +191,19 @@ lazy_static! {
pub static ref BASE_URL: String = env::var("BASE_URL")
.unwrap_or(format!("127.0.0.1:{}", env::var("ROCKET_PORT").unwrap_or(String::from("8000"))));
pub static ref DB_URL: String = env::var("DB_URL")
.unwrap_or(format!("postgres://plume:plume@localhost/{}", env::var("DB_NAME").unwrap_or(String::from("plume"))));
pub static ref USE_HTTPS: bool = env::var("USE_HTTPS").map(|val| val == "1").unwrap_or(true);
}
#[cfg(all(feature = "postgres", not(feature = "sqlite")))]
lazy_static! {
pub static ref DATABASE_URL: String = env::var("DATABASE_URL").unwrap_or(String::from("postgres://plume:plume@localhost/plume"));
}
#[cfg(all(feature = "sqlite", not(feature = "postgres")))]
lazy_static! {
pub static ref DATABASE_URL: String = env::var("DATABASE_URL").unwrap_or(String::from("plume.sqlite"));
}
pub fn ap_url(url: String) -> String {
let scheme = if *USE_HTTPS {
"https"

View file

@ -38,7 +38,7 @@ use diesel::r2d2::ConnectionManager;
use rocket::State;
use rocket_contrib::Template;
use rocket_csrf::CsrfFairingBuilder;
use plume_models::{DB_URL, Connection, db_conn::DbPool};
use plume_models::{DATABASE_URL, Connection, db_conn::DbPool};
use workerpool::{Pool, thunk::ThunkWorker};
mod api;
@ -51,7 +51,7 @@ type Worker<'a> = State<'a, Pool<ThunkWorker<()>>>;
fn init_pool() -> Option<DbPool> {
dotenv::dotenv().ok();
let manager = ConnectionManager::<Connection>::new(DB_URL.as_str());
let manager = ConnectionManager::<Connection>::new(DATABASE_URL.as_str());
DbPool::new(manager).ok()
}